Page 1
IndoorIndoor Air Quality in K-12 SchoolsLayered Risk (Dose) Reduction Amidst COVID-19
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Richard L. Corsi, Ph.D., P.E.H. Chik M. Erzurumlu Dean
Maseeh College of Engineering & Computer Science
Portland State University
@CorsIAQwww.corsiaq.com
Page 2
Schools & School Environments Matter
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• 1 in 5 Americans in schools each workday
• ≈ 58 M students (K-12: 1.6 to 1.8 yr inside schools)
• Mental, social, physical development
• Performance, illness, absence
• ≈ 3.7 M teachers + similar support staff
• Teachers; elevated work-related respiratory problems
Tak, S., et al. Journal of School Health, 2011.
• > $13,000/student-year US Census Bureau (2020), NEA (2020), educationdata.org
Page 3
Some Fundamentals
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Page 4
Sources of Emissions
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Breathing
• Speaking
• Singing
• Coughing
• Flushing?
• Resuspending?
• Virus not naked (embedded in particles)
• Particles = combo of mucous & saliva
• Small fraction of viruses infectious
Page 5
Particles & Viruses
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Virus ≈ 0.12 μm diameter
• Embedded in particles
• Emitted particles (< 0.3 - 200 μm)
• Particle diameter > virus
• Particle volume >> virus
•Vpart/Vvirus
• 1 μm 600 x
• 2.5 μm 9,000 x
• 10 μm 580,000 x
Particle size important
• Deposition onto indoor surfaces
• Removal in filters / masks
• Deposition in respiratory system
• How much and where
Page 6
Exposure Pathways & Fate
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
1. Direct contact
2. Fomites
3. Close contact: Near-field aerosols + droplets
4. Far-field aerosols
12
34
Page 7
RH %
Inactivation of SARS-CoV-2 in Aerosol Particles
van Doremalen, et al., NEJM, March 17, 2020
• Inactivation rate in aerosol particles
• < ventilation + filtration + deposition
• Assume no inactivation (safety factor)
> Fears, A.C., et al.
Morris, D.H., et al., bioRxiv, posted
October 16, 2020.
https://doi.org/10.1101/2020.10.16.341883
• Lower RH = shift to smaller particles
• Less deposition to indoor surfaces
• Deeper into respiratory system
Page 8
Inhaled Deposited Dose
Doseinhal,i = Ci (#/L) x B (L/min) x t (min) x fdep,i
• Ci = concentration of particles of size i
• emissions; mask; ventilation; control; deposition
• time infector is in space
• B = Respiratory minute volume
• activity (can vary significantly)
• t = Time in space with an infector
• fdep,i = Deposition of particles of size i in resp
• particle size; breathing mode; activity; (location)
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Convert # deposited to volume
Page 9
Layered Risk (Dose) Reduction Strategy (LRRS)
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• LRRS can lead to dose reduction > 95%
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Page 10
Reduce Source
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
“If there is a pile of manure in a space, do not try to remove the odor by ventilation.
Remove the pile of manure.” - Max von Pettenkofer (1858)
• Test & isolate
• Require masks (for all)
• De-densify (less occupants; innovate)
• Eliminate certain activities (singing, aerobics)
• Reduce speaking to extent possible
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Doseinhal,i = Ci (#/L) x B (L/min) x t (min) x fdep,i
Page 11
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Asadi, S. et al. Scientific Reports, 9:2348 (2019) doi.org/10.1038/s41598-019-38808-z
• Breathing ≈ order of magnitude lower than average speaking
Reduce Source: Speaking
Page 12
Possible Source: Resuspension of Particles
Students enter/exit
Lunch
Unoccupied
Ren, J. et al. Building & Environment (accepted)
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Re-suspension as source: VCT < Carpet
Page 13
Require Masks
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Universal mask wearing to capture infector
• Dual benefits
• 30% (I) & 30% (R) = 51% dose reduction
• 60% x 60% = 84% risk reduction
Problem = all masks off, e.g., lunch
• Outdoors if possible
• Quiet lunch (only teacher speaks)
• Rotating pods (teams) for mask off
• Mask down, eat, Mask up, next team up!
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Doseinhal,i = Ci (#/L) x B (L/min) x t (min) x fdep,i
Page 14
Cloth Mask Performance
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
http://jv.colostate.edu/masktesting/
Drs. John Volcken & Christian L'Orange
• Performance = strong function of material(s) & fit
• Particle size dependent
• Nice resource
• Select materials (includes data on breathability),
etc.
Page 15
Distance from Source (everyone)
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
50 -100 μm particles can travel > 6 ft (jet)
Horizontal distance traveled to settle 1.5 m
At free-stream air speed of 5 cm/s
dp (μm) t (1.5 m) x (m)
0.5 56 hr 100001 14 hr 2500
5 33 min 100
10 8 min 25
20 2 min 6
50 20 sec 1
Distancing?
• With masks
• Without masks
• Age / grades
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Doseinhal,i = Ci (#/L) x B (L/min) x t (min) x fdep,i
Page 16
Ventilate
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
https://www.nytimes.com/2020/07/17/
nyregion/coronavirus-nyc-schools-
reopening-outdoors.html
• Best = outdoors
• Mechanical (controlled)
• Natural (design openings)
• Infiltration
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Doseinhal,i = Ci (#/L) x B (L/min) x t (min) x fdep,i
Page 17
Ventilate
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
ASHRAE 62.1- 2019 Ventilation for Acceptable Indoor Air Quality (Pre-COVID)
5 L/s-person; 0.6 L/s-m2
If 24 students + 1 teacher in 60 m2 classroom = 5 x 25 + 0.6 x 60 = 161 L/s
161 L/s = 576 m3/hr; AER = 576 m3/hr / (60 m2 x 2.8 m) = 3.4/hr
The following modifications to building HVAC system operation should be considered:
• Increase outdoor air ventilation (disable demand-controlled ventilation and open out-
door air dampers to 100% as indoor and outdoor conditions permit).
• Additional recommendations on filtration, portable air cleaners, UVGI, T & RH, etc.
https://www.ashrae.org/file%20library/about/position%20documents/pd_infectiousaerosols_2020.pdf
ASHRAE Position Document on Infectious Aerosols
Approved by ASHRAE Board of Directors - April 14, 2020
Page 18
Ventilate
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Many schools under-ventilated or inappropriately ventilated
Absenteeism (Simons et al., Am. J. Public Health, 2010)
• Association: under-ventilation & absenteeism
• Strongest association: young students
Performance (Haverinen-Shaughnessy et al., Indoor Air, 2011)
• 100 southwestern schools/classrooms
• 87% w/ less ventilation than ASHRAE 62.1
• Each 1 L/s-student increase in ventilation:
• 2.9% increase math; 2.7% read
Ventilation matters (COVID-19 or not)
Page 19
Air Exchange Rates: Central Texas High Schools
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Permanent classrooms severely under-ventilated (Median < ½ ASHRAE 62.1)
• Generally higher ventilation in portable classrooms (but high variability)
• Portable classrooms – directly connected to outdoors
• Portable classrooms – more natural ventilation opportunities + infiltration
Lesnick, L.A. et al., ASHRAE Transactions (2017)
Page 20
Carbon Dioxide as Surrogate
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Elevated CO2 = inadequate ventilation
• Accumulation of pollutants, body odors
• Productivity decrements
• Increased absences (e.g., Shendell et al., Indoor Air, 2004)
• ∆ 1,000 ppm = 0.5-0.9% decrease in annual average daily attendance
• Elevated rebreathed fraction
• Greater probability of respiratory infections
• Lower CO2 (or RF): lower occupancy; increased ventilation
RF = (CO2,in – CO2,out) / CO2,breath
Page 21
CO2: Cumulative Distributions
115 K-8 classrooms; all day sampling; two school districts
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 20 40 60 80 100
Less Than or Equal to (%)
CO
2 C
on
cen
trati
on
(p
pm
)
Peak CO2
Average CO2
A B C
RF = 0.13
RF = 0.075
Median average RF = 0.025 (2.5%); Median peak RF = 0.044 (4.4%)
< 15% with average RF < 0.01; < 5% with peak RF < 0.01
Page 22
Rebreathed Fraction
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Central Texas High Schools (Year 1)
Lesnick, L.A. et al., ASHRAE Transactions (2017)
Median RF = 0.025 to 0.027 (2.5 to 2.7%)
Similar to previous K-8 results
Page 23
Estimates: Probability of Infection
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Rudnick-Milton model w/ 1 infector (m = adjusted for masks = 64% dual effectiveness)
Quanta generation rate: 67/hr for influenza; 135/hr for SARS-CoV-2
0.0
0.1
0.2
0.3
0.4
0.5
0.6P
robabili
ty o
f In
fectio
n (
fra
ctio
n)
Rebreathed Fraction
S2
Inf
S2 (m)
Inf (m)
CO2 = 698 ppm
S2 = 12% to 4.6% (m)CO2 = 1130 ppm
S2 = 28% to 11% (m)
Page 24
Filter
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Doseinhal,i = Ci (#/L) x B (L/min) x t (min) x fdep,i
“Improve central air and other HVAC filtration to MERV-13
or the highest level achievable.”ASHRAE Position Document on Infectious Aerosols (2020)
Kowalski & Bahnfleth (2002)https://www.researchgate.net/figure/Composite-of-all-MERV-filter-
models-based-on-initial-conditions_fig3_237558312
MERV: Minimum Efficiency Reporting Value
• Theoretical
• Can be worse
• System problems?
Page 25
Theory & Lab ≠ Practice
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Important to inspect for by-passCourtesy of Dr. Jeffrey Siegel, U Toronto
Courtesy of Dr. Atila Novoselac, UT Austin
(not a MERV 13 or 14)
MERV 14 filter, 20 Residences
Li and Siegel, Indoor Air (2020)
Page 26
Portable Air Cleaner (PAC)
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Proven: HEPA-based portable air cleaner
• High Efficiency Particulate Air
• Key: Clean Air Delivery Rate (CADR)
• CADR = η x Q
• η = single pass removal fraction (-)
• Q = volumetric flowrate (ft3/min)
• Example: η = 0.5; Q = 500 ft3/min
• CADR = 250 ft3/min
Shaughnessy, R.J., and Sextro, R.G., J of Occupational and Environmental Hygiene, 3: 169–181(2006)
IG
EPA.gov
Page 27
Portable Air Cleaner (PAC)
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Equivalent air changes per hour = EqACH = CADR/V
• Example: V = 600 ft2 x 8 ft = 4,800 ft3
• CADR = 300 ft3/min
• EqACH = 300 ft3/min/4,800 ft3 = 0.0625/min (or x 60 = 3.8/hr)
If λ = 2/hr
2 + 3.8 = 5.8/hr
66% reduction
Add to 64% masks = 88%!At steady-state
Page 28
Filter Microbiomes
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Filters have microbiomes (e.g., fungi growth on filter cake)
• Respiratory viruses have been found on filters
• Take precautions when changing filters (central or PAC)
• Do not agitate
• Mask / goggles
• Gloves / hand hygiene
• Bag it
Page 29
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Dissertation: Dr. Clive (Matt) Ernest, UT Austin
Disinfect (Air & Surfaces)
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Air: UVGI (can be very effective if done right)
Surfaces (wide range): residual, reaction by-products, worker exposure
Work-related asthma assoc w/ cleaning products
Rosenman et al., J. Occup. & Environ. Medicine (2003)
Page 30
Make Use of Time
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
0
30
60
90
120
150
180
1 2 3 4 5 6 7 8 9 10
Tim
e to
95
% D
ecay
(m
in)
B (hr-1) = AER + (CADR + hQr)/V
t95% = 3/B
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Doseinhal,i = Ci (#/L) x B (L/min) x t (min) x fdep,i
• Reduce continuous time indoors
• Reduce time w/ mask down at lunch
• Outdoor calm time after physical activity
• Classroom particle decay periods
Time for 95% decay of
aerosol particles
Page 31
Educate
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Entire school community
• Admin, teachers, staff, students, parents
• Target modes of communication
• People absorb differently
• English & Spanish
• Make use of existing tools – explore & educate
• Slides added to end of presentation
Require masks indoors
Make Use of Time
Reduce source
Require masks indoors
Filter
Distance from source
Ventilate
Educate
Disinfect (air & surfaces)
Page 32
SAFE AIR SPACES COVID-19 Risk Estimator
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Educational tool (layered risk reduction)
• Respiratory deposition & risk
• Factors: emissions, surface deposition
ventilation, filtration, masks
time in space, area & height
• Single zone (multiple coming)
• Far-field (working on near-field)
• Adaptable
Joint effort between
U of Oregon & Portland State
www.safeairspaces.com
Page 33
Scenario 1 – No Masks & Under-Ventilated
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
No masks; < ASHRAE 62.1; No filtration; High emitter; 2.5 hr exposure
Page 34
Scenario 2 – Masks & Under-Ventilated
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Masks for all; < ASHRAE 62.1; No filtration; High emitter; 2.5 hr exposure
Risk Reduction = 62%
Page 35
Scenario 3 – Masks + Increased Ventilation
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Masks for all; > ASHRAE 62.1; No filtration; High emitter; 2.5 hr exposure
Risk Reduction = 82%
Page 36
Scenario 4 – Masks + Increased Ventilation + Filtration + Outdoor Mask Break (20 min)
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Masks for all; > ASHRAE 62.1; Filtration; High emitter; 2.5 hr exposure
Risk Reduction = 92%
Page 37
Scenario 5 – 30 Minute Lunch w/o Masks
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Lunch in classroom: No Masks; > ASHRAE 62.1; No Filtration; High emitter; 0.5 hr
Relatively low risk in far field; near field (close contact) likely larger risk for scenario
Page 38
Closure
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• Schools critical for EVERYBODY
• Multiple benefits for children
• Need schools to be safe as possible
• Layered dose (risk) reduction works
• Use tools to educate and plan*
* Understand that estimation tools are valuable to show
trends, relative risks, & not necessarily exact numbers
Page 39
Acknowledgements
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
• International Society of Indoor Air Quality And Climate (ISIAQ)
• American Industrial Hygiene Association (AIHA)
• United States Environmental Protection Agency (USEPA)
Modeling partners:
Kevin Van Den Wymelenberg & Hooman Parhizkar (both U Oregon)
Additional colleagues who provided information for this presentation:
Jose Jimenez (CU Boulder), Atila Novoselac (UT Austin), Jeffrey Siegel (U Toronto)
Ex-students who did the work to generate data used in this presentation:
Matt Ernest, Sangeetha Kumar, Leigh Lesnick (all UT Austin)
My Executive Assistant for help with logistics: Brandi Cobb (Portland State)
Page 40
Some Additional Resources & Tools
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Page 41
ASHRAE Epidemic Task Force - Schools
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
https://www.ashrae.org/file%20library/technical%20resources/covid-19/ashrae-reopening-
schools-and-universities-c19-guidance.pdf
Page 42
EPA Tools for Schools, etc.
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
https://www.epa.gov/iaq-schools
Page 43
Harvard T.H. Chan School of Public Health
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
https://schools.forhealth.org/
Page 44
AIHA – Reopening Guidance
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
https://aiha-assets.sfo2.digitaloceanspaces.com/AIHA/resources/Reopening-
Guidance-for-Schools-K-12_GuidanceDocument.pdf
Page 45
FATIMA Model (NIST)
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
https://www.nist.gov/services-resources/software/fatima
Page 46
CU Boulder Aerosol Transmission Estimator
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University
Courtesy Jose L. Jimenez
Page 47
Aerosol Science & Indoor Air Researchers
Richard L. Corsi, Ph.D., PE.
Dean, Maseeh College of Engineering & Computer Science, Portland State University